Finding the right attorney is a routing problem, not a search problem. A list of every family lawyer in your county does not tell you which one handles your specific situation well. Attorney Match on LawSensai treats the question as routing. It uses the ANS Core engine to rank attorneys against your actual matter signals, then warm-introduces you to the ones most likely to accept and handle the work.
This post explains how the ranking is built, what signals the ANS Core uses, and what happens after you submit a matter.
What Attorney Match does
Attorney Match takes the facts already in your LawSensai matter and runs them through the ANS Core. The Core scores attorneys in our network against your facts and returns a ranked shortlist. From there you can request a warm intro to the top candidates. The attorney either accepts the intro or passes, and the Core learns from that signal for the next routing decision.
Nothing about your matter is sent to an attorney without your explicit consent. The shortlist is private to you until you press request.
The signals the ANS Core uses
The Core ranks attorneys on five primary signals:
Practice area. Criminal defense, family law, personal injury, small claims, immigration, employment, and the long tail of subspecialties under each. An attorney who handles divorce mediation is not the same as one who handles contested custody trials, and the Core treats those as distinct.
State. Bar admission is by state. The Core never routes a matter to an attorney who is not admitted in the state where the matter sits. This is a hard filter, not a soft preference.
County. Local court familiarity matters. A criminal matter in one county can have different prosecutor norms, judge tendencies, and pretrial release practices than the next county over. The Core prefers attorneys with documented county experience.
Charge or claim type. The specific charge code on a criminal matter, the cause of action on a civil matter, the custody-versus-support split on a family matter. The Core uses the matter taxonomy to match attorneys whose history shows that exact pattern.
Per-cohort federated bandit signals. The Core runs a multi-armed bandit per cohort of similar matters. Cohorts are defined by state, county, and matter type. Within a cohort, the bandit tracks which attorneys accepted, which declined, which closed the matter, and which were rated highly by clients. Those signals feed the next ranking. The bandit is federated so cohort-level learning does not leak individual matter facts across cohorts.
What happens after you submit a matter
The flow goes intake, score, shortlist, warm intro, accept or pass, next attorney if needed. The Core does not blast your matter to a hundred attorneys. It contacts them one at a time or in a small batch, depending on the urgency signal pulled from the Court Date Tracker. If the hearing is in 72 hours, the urgency is high and the batch is wider. If the matter has no near-term deadline, the Core contacts the top candidate first.
Every routing decision is recorded in brain_decisions. The audit-log hash chain captures the matter fingerprint, the ranked list, the contact order, and the accept or pass outcome. Support can reconstruct any routing decision after the fact.
What you see as the user
You see the shortlist with three to five attorney profiles. Each profile shows the attorney's practice area focus, the state and county they cover, and a short summary of why the Core ranked them for your matter. You do not see internal scores. You do see the bandit signals expressed as plain language, for example documented experience with this charge type in this county.
You pick whether to send the warm intro. The attorney sees a redacted summary of the matter. They accept or pass within a set window. If they pass, the Core moves to the next attorney on the shortlist.
Public Defender fallback
The Core has 51 Public Defender offices seeded for the matters where a PD referral is appropriate. If no private attorney accepts and the matter posture supports a PD referral, the Core falls back to the appropriate PD office and warm-introduces you there. We cover the full PD warm-intro flow in a separate post.
Spanish language support
Attorney Match runs in Spanish. The matter intake, the shortlist, and the warm intro request page are localized. The shortlist will surface attorneys whose profile indicates Spanish-language practice when your matter is created in Spanish.
Safety and audit
Every decision the ANS Core makes is recorded. The brain_decisions table holds the ranked list, the score inputs, and the outcome. The aggregate decision counts and safety findings are published on the Trust Center at lawsens.ai/trust/brain. If a routing decision ever looks wrong, support can pull the exact decision record and review it.
No attorney pays for placement. The Core ranks on the signals above, not on payment.
Common misreads we see new users make
Misread one: thinking the shortlist is a referral list. It is a ranking against your matter. The attorneys on it have been selected because their history matches your matter taxonomy, not because they paid to appear.
Misread two: assuming the attorney sees your full matter. The warm intro shares a redacted summary. The attorney does not see your private documents, the Document Vault, or your Court Date Tracker until you explicitly grant access after the engagement is set up.
Misread three: treating a pass as a verdict on your matter. Attorneys pass for many reasons including caseload, conflict checks, and scheduling. A pass is a routing signal for the Core, not a judgment on your case.
Practical next steps
Step one: start at lawsens.ai/match. Pick the practice area and answer the short intake. The intake is the same one the Court Date Tracker uses, so any matter you already have is prefilled.
Step two: review the shortlist at lawsens.ai/match/results. Read the profile summaries and pick which warm intros to send.
Step three: if no private attorney accepts and your matter posture supports it, accept the PD warm intro the Core proposes at lawsens.ai/match/pd. The PD office handles the next steps from there.
How Attorney Match connects to the rest of LawSensai
Attorney Match is the connective tissue between the rest of the product and a human attorney. The Court Date Tracker tells the Core how urgent your matter is. The matter intake fields feed the practice area, state, county, and claim type into the ranking. After an attorney accepts, the Document Vault and the Court Date Tracker can be shared with the attorney through scoped access. Every step is logged in brain_decisions and published in aggregate on the Trust Center.
This post is informational and is not legal advice. Attorney Match is a routing engine. The attorney-client relationship begins when you and the attorney agree to engage.
Read more
- lawsens.ai/product/attorney-match
- lawsens.ai/trust/brain
- lawsens.ai/trust
- lawsens.ai/help/warm-intro
- lawsens.ai/help/public-defender-fallback
- American Bar Association lawyer referral guidance at americanbar.org
Last verified: 2026-04-09.